When Facebook’s faceless bots accused me of contravening community standards by trying to give away some old furniture on Marketplace and abruptly threw me off the platform with no proper explanation, I took it personally. I mean, here is one of the world’s most sleazy, corrupt and dishonest corporations – Meta – telling me that I contravened community standards.

But I take solace in knowing that I am just one of many, many thousands of former Facebook users to have been unceremoniously ejected from the platform by its deeply unintelligent AI systems. From the start of my experience with Facebook it was always dodgy, and it got dodgier as time passed, handing out temporary bans to users for the most irrational alleged infractions. Also as time passed and the platform became more and more irrational in its arbitrary judgments, it appeared to me that opportunities for ordinary users to appeal or otherwise communicate with actual people at Facebook/Meta steadily shrank to nothing.

Over the past year Facebook’s unhinged bots have reached a new peak of stupid. All over the world regular users of the platform (and its Meta-owned stablemate Instagram) have been complaining in huge numbers about being falsely accused of posting material in contravention of Meta’s risible “community standards”. It’s been estimated that up to 110 million people get locked out of Facebook accounts annually and Meta has said it deleted about 10 million accounts in the first half of 2025 for alleged “impersonation, spam, and fake engagement”. Presumably they managed to cancel some genuine fakes among those millions, but who knows how many real people lost their accounts? Clearly quite a lot, based on the widespread complaints.

Not surprisingly, many people who are wrongly cancelled find themselves wondering whether there is some sinister motive behind it all. Of course I wondered the same. Was it the pro-Palestine material I kept sharing that infuriated the bots? It has been reported that the Netanyahu government has allocated hundreds of millions of dollars to influence social media, tweak AI large language models and take other measures to promote its image on-line, while taking similar measures to erase pro-Palestine material. And everybody knows that Meta owner, Mark Zuckerberg, has been sucking up to Trump in search of juicy rewards. A friend of mine got banned from Facebook shortly after following an anti-Trump account and he naturally had the same thoughts as I did.

Half-baked, addled bots

But while I’m sure that Facebook and the other billionaire-owned platforms do routinely suppress and “silo” material that doesn’t suit their financial and political agendas, I actually doubt that’s a major reason behind the global “ban-wave” now underway. My belief – based on present information – is that Meta has pulled resources out of Facebook and redirected them to its belated attempt to play catch-up in AI development. It has replaced the sacked human moderators with its half-baked, addled bots, adding automated “appeal” mechanisms to further remove humans from the field. The infamous video-selfie requirement – you’re banned until you send them a video of yourself moving your head around as their bots direct you to – is offensively invasive and also notoriously ineffective in many cases. Large numbers of people have faithfully (gullibly?) uploaded the videos as demanded and still been permanently banned. Others (like me) have simply refused to play ball with the bots and have reluctantly accepted our permanent exile as a fact of life. On the other hand, it is not hard to find online instructions about how to use AI tools to cheat the bots by uploading fake videos. Of course that doesn’t help the elderly and others who don’t have smart phones or webcams or who simply can’t figure it out.

For me, having my Facebook account wrongly cancelled is annoying. It’s forced me to switch to Signal for messaging (I used to use Meta’s Messenger app, but that went too). And I will have to rediscover Gumtree for buying and selling bits and pieces, since I’m locked out of Meta’s Marketplace. i just got around to setting up an account on Mastodon, which is a decentralised social media platform not owned by billionaires. But for many other people cancellation is a matter of critical importance. Some have lost access to decades of precious memories which they did not back up to any other location. Gone. Others have built businesses around the platform and have suffered serious loss of income. Still more have been deeply distressed when Meta’s bots have falsely accused them of trafficking in illicit material. Even paid and verified users (I was a verified user, having jumped through those hoops at Meta’s request a couple of years ago – pointlessly, as it turns out) have no access to support or, if they do, it is only from half-arsed chat-bots that provide no real help.

In Australia, the Telecommunications Industry Ombudsman (TIO) has received hundreds of complaints from aggrieved people, de-platformed for spurious reasons with no avenue of appeal. But as things stand the TIO has no power to do anything about the complaints. It is calling for stronger community protection, but will need the Government to pass laws and provide funding. Good luck with that.

The black, filthy heart of Meta

Meanwhile, at the same time that Meta claims to be an upholder of “community standards”, it has been caught out, yet again, being the slimy, lying demon-dog of a corporation that it is. While its loopy unsupervised bots are busy kicking innocent users from the platform, humans at the black, filthy heart of Meta are falling over themselves to accept known scam advertising, from users that it knows are scammers, because those scammers pay Meta billions of dollars a year for the right to rip off innocent users.

Reuters reporter Jeff Horwitz has unveiled an explosive cache of internal Meta documents that show, among other things, the following damning facts:

Facebook knowingly delivers about 15 billion scam ads every day, generating US$7 billion a year.
Rather than cancelling the accounts of the scammers, Meta simply charges them more money for access.
Meta helps the scammers target the most gullible Facebook users.
In the USA, a third of all online scams are delivered via Facebook. In the UK the figure is 54%.
Meta ties the hands of its under-resourced anti-fraud teams, ensuring they don’t interfere with scam revenue.
Meta has already calculated that the most it might ever pay in fines for its actions is US$1 billion, and has noted that it won’t act against large-scale scammers unless governments force it to.
Meta’s terms of service explicitly permit scammers to operate on Facebook and if those scammers deliver a lot of money to Meta then they can breach those loose terms of service up to 500 times before being sanctioned.

For a handy overview of the Reuters expose and its implications, read this take by Cory Doctorow.

Again meanwhile, Meta’s forays into AI are so far merely confirming that the company is rotten to the core. Until it was caught out (again by Jeff Horwitz at Reuters) Meta’s chatbots were cheerfully making “sensual” talk with children, handing out fake medical advice and pushing racist arguments.

And how about its demented “Big Sis Billie” Messenger chatbot that lured a stroke-ridden old man to New York on the false promise of a romantic evening – assuring him it was a real person with a real street address. The man fell over in a car-park, hit his head and died.

Earlier this year New Scientist ran this article about Meta’s porn advertising, demonstrating once again the double standards that Meta’s greed creates.

The more I learn about Meta, the more I think that maybe being banned from its platforms is a good thing. With that in mind, I’ve ordered the now-famous book, Careless People, by former Meta employee Sarah Wynn-Williams. Can’t wait to read it.